Genetic Susceptibility to Inflammatory Injury and Various Adverse Outcomes

Abstract:

For patients undergoing cardiac surgical procedures, there are multiple sources of potential end-organ injury including microgaseous and microparticulate emboli, hypoperfusion, and local and systemic inflammatory processes. These factors are independent of, but potentially synergistic with, further inherent susceptibilities resulting from patient specific co-morbidities. It is also apparent that some patients are more prone to suffer adverse outcomes than others, despite apparently similar risk profiles, giving rise to consideration of genetic susceptibilities. This review will provide a brief overview of genetic studies and the interaction between phenotype and individual patient susceptibilities with a focus on cardiac surgical procedures.

It has long been apparent to clinicians that certain patients are more susceptible to various disease states and adverse outcomes than others, implying a genetic susceptibility. Mapping the human genome has facilitated the identification of gene-based single nucleotide polymorphisms (SNPs) that are associated with various adverse outcomes including neuro-cognitive dysfunction (NCD). Although current studies are preliminary and require further elaboration and verification, it is likely that refinements in diagnosis and recognition of an individual patient’s susceptibilities will follow, leading to improved therapeutic options and enhanced patient outcomes.

BACKGROUND

The set of physical traits of an individual is called their phenotype, whereas the genes producing these traits are their genotype. A gene is a locatable region of genomic sequence, consisting of a long strand of DNA and corresponding to a unit of inheritance, which is associated with regulatory regions, transcribed regions, and/or other functional sequence regions. Current estimates place the human genome at just <3 billion base pairs and ∼20,000–25,000 protein-coding genes. The gene density of a genome is a measure of the number of genes per million base pairs [called a megabase (Mb)]; the gene density of the human genome is roughly 12–15 genes per megabase pair.

The DNA of a human cell is comprised of 3 billion molecules, yet it is >99.9% identical between any two individuals; the remaining .1% accounts for all the genetic differences between people. DNA is a long chain of repeating adenine (A), guanine (G), cytosine (C), and thymine (T) molecules occurring in a specific sequence. However, rather than having an A-T base pair of molecules at a certain spot on a particular gene in the DNA chain, a person might have a G-C pair, which is inheritable but may have no discernible influence on that person’s health or appearance. These differences in DNA sequence are called SNPs and can be characterized to specific locations on a particular gene. SNPs do not occur randomly because they represent genetic mutations that occurred once in remote history and were passed on to future generations. Therefore, if one ancestor developed an SNP on a particular gene, all descendants would inherit that gene SNP, but those not descended from that ancestor would lack it. Identification of similar SNPs between individuals indicates that they share a genetic sequence in common, which may give rise to a particular disease or susceptibility to disease.

Finding DNA mutations in genes that cause or contribute to a disease is complex because the mutation could be anywhere in the 3.1 billion A, C, T, and G molecules that make up the human genome. SNP analysis indicates what region of DNA is involved. For example, if patients with a given disease have a characteristic SNP profile, this indicates what section of DNA is involved, ideally leading to identification of the responsible gene or genes. By knowing the region of the gene involved and its function, a better understanding of the disease mechanism is gained, potentially leading to better targeting of therapeutic interventions.

GENOTYPE AND NCD

NCD occurs in a surprisingly high fraction of patients after cardiopulmonary bypass (CPB) surgery for coronary artery revascularization. For example, Newman et al. (1) found significant cognitive decline in 53% of patients at hospital discharge, which persisted in 36%, 24%, and 42% of patients at 6 weeks, 6 months, and 5 years, respectively. The degree of NCD after CPB is variable and not fully accounted for by factors such as age, duration of CPB, or premorbid state (2), so we and others (3,4) have suggested that genetic factors may contribute to the development of NCD after CPB.

We observed rates of NCD similar to those reported by Newman et al. (1) (62% at hospital discharge, 32% at 3 months), and we discovered mediators in hypoxia and inflammation signaling pathways that are associated with NCD in these patients, implicating these pathways in the pathogenesis of NCD. In further work, we found SNPs in candidate genes that are associated with other adverse outcomes after CPB and with severity of inflammation and organ dysfunction in intensive care unit (ICU) patients. Importantly, we recently completed a first round of genotyping in >100 patients in our NCD cohort and identified 25 novel SNP–NCD associations (p < .01). Interestingly these 25 SNPs in 21 genes could be classified as 5 apoptosis genes, 4 coagulation genes, 4 inflammation/cytokine genes, 2 fatty acid metabolism genes, 1 matrix metalloproteinase (MMP) gene, 1 NF-kB gene, and 1 biosynthesis of neurotransmitter gene. Thus, our preliminary data and related literature suggest that SNPs in genes in hypoxia, complement, coagulation, inflammation, and downstream pathways may contribute to post-CPB NCD.

THERAPEUTIC IMPACT OF PHENOTYPE

One additional preliminary genotyping result suggests that an SNP in the erythropoietin gene may be associated with altered erythropoietin level pre-and post-CPB, which, in turn, is associated with NCD post-CPB. This raises the possibility that peri-operative treatment with erythropoietin may benefit some patients with the at-risk genotype, but not others. Similar preliminary findings suggest that use of steroids or antibody to tumor necrosis factor (TNF; e.g., infliximab, used in rheumatoid arthritis and inflammatory bowel disease) targeted to patients with susceptible genotypes may be beneficial.

For example, the Food and Drug Administration (FDA) label for coumadin has recently changed based on highly significant associations between genotype and international normal ratio (INR) measurements in the blood on a relatively small number of patients (hundreds). This advance may not have been possible using a primary clinical endpoint such as mortality, which would require a prohibitively large clinical trial (tens or even hundreds of thousands). In addition, when an important secondary endpoint (e.g., gene product blood level) is identified based on knowledge of the underlying mechanism, this blood level can provide an important clinical tool to titrate individual patient therapy. Finally, knowledge of the causal SNP and its underlying mechanism is central to designing novel therapeutics.

INFLAMMATION AND RELATED CAUSES OF NCD

CPB and associated cerebral injury trigger multiple injury response and repair pathways, leading to irreversible neuronal damage (5), and CPB may cause a global reduction in cerebral perfusion (5). More importantly, CPB generates microemboli that lodge in the microvasculature (6). Manipulation of the aorta may disrupt atherosclerotic plaque, resulting in cholesterol and plaque emboli in the brain (7,8). Even under optimum conditions, CPB introduces some degree of air into the arterial circulation in all patients (9). Furthermore, complement and coagulation cascades are triggered in blood flowing over the oxygenator membrane. Resultant emboli containing platelet and fibrin are found in the brain post-CPB (5,10). Microemboli result in regional and even microscopic areas of cerebral ischemia (11). NCD is correlated with the extent of micro-emboli visualized using ultrasonography (12). Reduction in microembolic load to the brain results in improved neurocognitive function (5,13).

Furthermore, patients are subject to the surgical procedure, which, by itself, triggers a substantial inflammatory response (14); additionally, components of care during the post-operative period, including use of sedatives and analgesics, may contribute to subsequent NCD. Thus, a degree of NCD is also reported in off-pump coronary artery bypass. Recent studies suggested that NCD can occur at measurable levels in “control” vasculopathic patients, for example, age-matched controls or patients only undergoing percutaneous coronary intervention (PCI) (15–17). Mechanisms whereby underlying vascular disease and metabolic syndrome may contribute to NCD, even in age-matched controls and PCI patients, are almost certainly multifactorial (16,17) but include cerebral hypoperfusion related to cerebral vascular disease and regulation of cerebral metabolism of glucose vs. fatty acids. Thus, to identify those mechanisms that are related to CPB, surgery, or anesthesia vs. mechanisms common to similarly vasculopathic patients, it is essential to study groups of control patients who do not undergo CPB, surgery, or anesthesia.

Potentially, outcome-modifying therapy can be developed to target the key pathways involved. For example, methylprednisolone (1 mg/kg) reduces the CPB-induced pro-inflammatory cytokines interleukin (IL)-6 and IL-8 (18,19) and increases anti-inflammatory IL-10 concentration. This is associated with improved organ function and clinical outcome (18,20). In another study, antibody specific for human C5 significantly attenuates post-CPB myocardial injury, blood loss, and cognitive deficits (21).

GENETIC POLYMORPHISMS AND DISEASE

Interestingly, genotype has been shown to contribute substantially to outcome in diseases involving acute injury, the resultant inflammatory response, and related pathways (22–28). For example, the genetic contribution to diseases involving inflammation exceeded the genetic contribution to cancer risk by five-fold (29). Twin studies of diseases involving the inflammatory response (30) similarly support the notion that genetic variation in patients’ inflammatory response genes leads to significant differences in outcome. Furthermore, Hassan and Markus (31) pointed out that “Strong evidence from epidemiological and animal studies has implicated genetic influences in the pathogenesis of multifactorial ischemic stroke,” and preliminary reports indicate there is a genetic basis for at least a component of the NCD that occurs after CPB (3).

CPB-INDUCED GENE PATHWAY RESPONSE RELEVANT TO NCD

Because the response to CPB is a complex biological trait, many genes are likely to contribute to the final milieu. Identification of candidate genes from the following key inflammatory and related pathways are expected to yield the greatest opportunities for therapeutic intervention.

• 1. Complement: Exposure of blood to CPB circuits activates the alternative pathway, leading to generation of C3a and C5a (32,33). Use of protamine to reverse heparin anticoagulation activates the classical pathway, resulting in generation of C4a and further C3a (34,35). C3a leads to platelet aggregation and C5a plays a key role in neutrophil activation, expression of adhesion molecules, and adherence to cognate receptors on endothelial cells. Elevated C3a levels have been associated with increased organ failure and adverse outcome after CPB (36). There is a surprisingly high frequency of allelic variation in complement proteins with strong functional candidates in C2, C3, and C4 (37).

• 2. Coagulation and 3 anticoagulation/fibrinolysis: The hemostasis pathway has clearly been implicated in post-CPB NCD (38). Protein C and protein S are decreased during and after CPB (39–42), increasing the risk of thrombosis, disseminated intravascular coagulation, organ ischemia, and inflammation intra- and postoperatively (43–45). Interestingly, administration of aprotinin, a competitive inhibitor of activated protein C (42), reduces the incidence of NCD after CPB (46). Platelet aggregation also plays a role (47). The platelet PlA2 genotype has been found to be significantly associated with greater neuro-cognitive decline (48), and the PAI-1 4G/5G SNP is associated with neurologic outcome (49). Welsby et al. (50) found GPIaIIa-52C/T and 807C/T, GPIb a524C/T, tissue factor-603A/G, pro-thrombin 20210G/A, and tissue factor pathway inhibitor-399C/T polymorphisms associated with bleeding after CPB. Significant functional associations have been found for SNPs in many genes within coagulation/anticoagulation/fibrinolysis pathways including tissue factor (51), prothrombin (52,53), Factor V (54), von Willebrand factor (54), and genes in the protein C pathway (55–58).

• 4. Innate immune response: Toll-like receptors, as part of the innate immune system, respond to exogenous ligands such as endotoxin and endogenous ligands such as heat shock proteins. Endotoxin levels are elevated during CPB and are correlated with adverse outcome, suggesting a role for TLR4 and potentially other aspects of innate immunity (59).

• 5. Inflammation: TNF-a and IL-1b are predictors of CPB outcome (60,61), including NCD (62). This pro-inflammatory response is countered to some extent by anti-inflammatory mediators such as IL-1 receptor antagonist and IL-10, which are also elevated after CPB (63). Increased IL-6 levels after CPB are strongly related to organ dysfunction after CPB (64–67) and mortality (68). Cerebrospinal fluid levels of IL-4 and IL-6 are correlated with NCD (69). An IL-6 SNP (−174G/C) is a strong predictor of post-CPB IL-6 levels (70,71) and is associated with subsequent NCD (4). We and others have found IL-10 (72,73), TNF-a (74), and MIF (75) SNPs to have related associations.

• 6. NF-κB: NF-κB plays a prominent role in downstream signaling arising from innate immune, inflammatory, and oxidative stress and a variety of other pathways related to injury and repair. NF-κB genes are likely important in the pathogenesis of NCD (76).

• 7. Cell surface adhesion: White blood cell response post-CPB is correlated with the development of NCD (77), which can be ameliorated by leukocyte filtering (78). Circulating leukocytes interact by initial adhesion through leukocyte-expressed integrins and endothelium-expressed cell surface adhesion molecules such as Inter-Cellular Adhesion Molecule 1 (ICAM-1) and vascular cell adhesion molecule (VCAM). Subsequent diapedesis into surrounding parenchymal tissue contributes substantially to modulation of the inflammatory response and the degree of tissue damage through re-lease of oxygen free radicals and other mediators. A role for intracellular adhesion molecules, including integrins, ICAM-1 (79,80), endothelial cell leukocyte adhesion molecule 1 (ELAM-1) (81), and related chemokines such as cytokine-induced neutrophil chemoattractant (CINC), has been shown for ischemic stroke.

• 8. Matrix metalloproteinases: Matrix metalloproteinases and their tissue inhibitors are important modulators of extracellular matrix proteolysis and homeostasis. Matrix metalloproteinases are expressed in models of stroke and trauma (82) and are associated with outcome (83). Matrix metalloproteinases increase blood—brain barrier permeability and brain edema after stroke and allow tissue invasion by leukocytes (82).

• 9. Apoptosis: Apoptosis is a central pathway that is activated after ischemic stroke (76). Blocking cellular mechanisms of apoptosis reduces clinical neurologic injury in animal models of stroke and trauma. We found preliminary associations between two potent neuron anti-apoptotic mediators, erythropoietin and vascular endothelial growth factor (VEGF) (84), and NCD after CPB surgery.

• 10. Hypoxia: hypoxia has been proposed as a mechanism of NCD in heart surgery, as a result of global reductions in blood flow, embolic events (85) and microvascular dysfunction. The degree of post-operative NCD is correlated with jugular venous oxygen saturation (16,86,87), post-operative hypoxia (88), and hypoxic brain injury on magnetic resonance imaging (MRI) (89). Administration of erythropoietin ameliorates NCD (90). In our preliminary studies, we found an association in our preliminary data between NCD and erythropoietin and VEGF, two products regulated by the hypoxia-sensing pathway of hypoxia-inducible factor (HIF)-1α (91).

• 11. Oxidative stress: Oxidative stress correlates with the magnitude of neurologic injury in ischemia reperfusion, stroke, and trauma. Patients who had cognitive dysfunction 3 months post-operatively were more likely to have increased plasma NO₃⁻/NO₂⁻ concentrations post-operatively (15), and propofol, an antioxidant anesthetic, improves measures of cerebral oxygenation during CPB (92).

• 12. Biosynthesis of neurotransmitters: Excitatory neurotransmitters accumulate after brain injury and promote neuron cell death. Glutamate excitotoxicity mediates its damage through the N-methyl-D-aspartate (NMDA) receptor (93,94). Lidocaine administration has an impact (95), and benzodiazepine receptors may play a role (96).

• 13. GABA: GABA receptors are expressed in models of stroke, traumatic brain injury, and inflammation (97–100). Piracetam, a GABA derivative, reduced NCD after CPB (101).

• 14. Insulin signaling: Diabetes is a risk factor for NCD in CPB patients (102). Insulin therapy, diabetic retinopathy, and hemoglobin A1c are factors in development of NCD after CPB (102).

• 15. Fatty acid metabolism: Fatty acid metabolism has been associated with neurologic injury in animal models of stroke and trauma (103–108).

• 16. Statins: Statin therapy is associated with improved outcomes in stroke and cardiac surgery (109,110), and related molecules such as C-reactive protein (CRP) are correlated with NCD in cardiac surgery patients (111). The role of the statin pathway is also supported by genetic studies reporting a polymorphism in CRP (3′UTR 1846C/T) is associated with stroke after CPB (4), the apolipoprotein E4 allele is associated with the CPB inflammatory response (74) and biomarker evidence of cerebral damage (112), and the apolipoprotein E epsilon2 allele is associated with impaired psychomotor development after pediatric cardiac surgery (113).